SHOET MEMOIRS ON METEOROLOGICAL SUBJECTS. 419 



&c., although it follows from the law of Mariotte and Gay Lussac that 



- has the value ^ , as you yourself* have used it, page 23. 



Furthermore, in the deduction of the formula of Peslin, the approxi- 



mate value of q. 0.G23 - , is used, whereas the introduction of its true 



value offers no difficulty. Considering these two circumstances, the 



expression for — receives a somewhat different form, which I will here 



deduce. 

 In the equation that applies to a change of condition of any gas, 



^ _ _ ET dp 

 dQ=cdt— -j-'-z » 



the left-hand side is if dry air changes its condition without addition 

 or subtraction of heat. But if it is moist saturated air that ascends 

 without receiving heat from the outside, then we can use the preceding 

 equation by putting for dQ, not 0, but that heat which is liberated by 

 the condensation of the elementary quantity dq of aqueous vapor. Let 

 the specific heat of moist air be c'; for moist air the constant E can very 

 approximately retain unchanged its former value 29.3 ; the equation 

 then becomes 



^ J P 



The left-hand member has the minus sign, because it represents a 

 positive quantity of heat, and dq is a negative quantity, namely, the 

 diminution of the quantity of vapor there present. To this we add 

 the equation 



(2) dp = p. dh, 



which expresses the diminution of pressure with altitude. In case that 

 the ascent of a moist current in dry air is considered, then for p the 

 specific gravity of dry air is to be taken, since in a narrow, slightly 

 rotating current the pressure can differ only extremely little from the 

 pressure at equal altitude outside. 

 Since now 



* I have, in fact, on page 28, put 



P_ 1 _ 1 



p BTo 29.3X273' 



whereas it should have been 



29.3 X (273 + t) ' 

 where t is the temperature in Celsius degrees. Since, however, t is always small in 

 comparison with 273, this assumption of To, instead of T, has only a very slight influ- 

 ence upon the result. The introduction of an approximate value for q materially sim- 

 plifies the numerical computations. — J. Hakx. 



